Some implantable sensors are sensitive to changes that take place within a cardiac and/or respiratory cycle. An example of such an implantable sensor is a venous oxygen saturation sensor (SvO2), which can be placed, e.g., inside a blood vessel or a chamber of a heart. An SvO2 sensor includes light sources to transmit light of multiple wavelengths. The transmitted light interacts with surrounding blood, and is either absorbed or reflected back and detected by one or more light detector of the SvO2 sensor. The portions of the light absorbed and reflected depend on the oxygen level of the blood. Thus, the levels of the light detected by the light detector(s) can be used to measure venous oxygen saturation levels. However, other factors can also affect how much light is absorbed and reflected. For example, the placement of the sensor can have a strong effect on how light is absorbed and reflected. More specifically, if the light sources face a vessel wall or chamber wall, the amount of reflected light might be greatly increased during heart contractions because light would reflect off of the vessel or chamber wall without interacting with the blood. In this case, the measurements made during contractions would likely be inaccurate. Direction of blood flow and cell orientation relative to the sensor can also influence sensor measurements by changing the balance between reflection and absorption. For another example, movement of the chest wall, due to respiration, can also alter sensor measurements.
Sensor measurements can be used in various manners, such as, for feedback when adjusting pacing parameters, for monitoring heart failure patients, for triggering therapy and/or alarms, etc. From the above description, it is clear that various factors can affect the accuracy of sensor measurements. It would be undesirable and potentially detrimental to use such inaccurate sensor measurements for feedback, monitoring, triggering therapy and/or alarms, or the like. Accordingly, there is a need to increase the accuracy of sensor measurements obtained from implantable sensors.